Management device, management method, and control system

Abstract

This management device (10) comprises: a data acquisition unit (11) that acquires operation data including time series data relating to the operation of a plant (20), and region-of-interest information representing a predetermined region in the data space of the operation data; a region-of-interest determination unit (14a) that determines whether the state of the plant (20) is within the region of interest on the basis of the acquired region-of-interest information; a state category map generation unit (14b) that uses the acquired operation data and the region-of-interest information as a basis to generate state category maps in which the state of the plant (20) is associated with a state category indicating a predetermined classification; and a state category calculation unit (14c) that calculates the state category of the corresponding plant (20) on the basis of the generated state category maps. The state category map generation unit (14b) classifies the operation data as inside or outside the region of interest and generates state category maps corresponding respectively to the interior and the exterior of the region of interest.

Description

Management device, management method, and control system

[0006]

[0001] The present disclosure relates to a management device, a management method, and a control system.

[0002] Regarding the management of a control target, for example, the technique described in Patent Document 1 is known. In this publication, in paragraph 0018, it is described that "FIG. 5 shows an example in which 15 categories are mapped onto a two-dimensional plane (XY plane). The mapping result onto the two-dimensional plane that makes it easier to grasp the positional relationship between categories is hereinafter referred to as 'category map data (map data)'. The category map data is used like a map in a navigation system. Here, the multidimensional scaling method (MDS: Multi Dimensional Scaling) is used for mapping. MDS is a method of determining coordinates so as to reproduce the distance information on a q-dimensional space when distance information is given for a plurality of p-dimensional data. When p>q, the distance information cannot necessarily be accurately reproduced, but it is effective for grasping the approximate positional relationship. In particular, when q = 2, it can be mapped onto a plane of data in a multidimensional space, so it can be used for visualizing multidimensional data."

[0003] Japanese Patent Application Laid-Open No. 2019-016214

[0004] The technique described in Patent Document 1 calculates the spatial distance of operation data of a control target (plant), extracts a plurality of candidates for target operation data, and presents a target for an operation state for appropriately operating the control target.

[0005] However, in the technique described in Patent Document 1, when the state of the plant changes due to a disturbance and the plant efficiency or the product yield decreases, the problem is how to appropriately set the target state of the target plant rather than improving the control characteristics up to the target value. Therefore, a plant operation support device presents a target for an operation state for appropriately operating the plant starting from the current operation state of the plant.

[0006] Therefore, the plant operation support device described in Patent Document 1 does not particularly mention appropriately improving the control characteristics.

[0007] Therefore, the present invention aims to provide a management device, a management method, and a control system that can appropriately improve the control characteristics of a controlled object.

[0008] To solve the aforementioned problems, the management device according to this disclosure includes: a data acquisition unit that acquires operation data including time-series data relating to the operation of a controlled object and attention area information representing a predetermined area in the data space of the operation data; a attention area determination unit that determines whether the state of the controlled object is within the attention area based on the acquired attention area information; a state category map generation unit that generates a state category map that associates the state of the controlled object with a state category indicating a predetermined division based on the acquired operation data and the attention area information; and a state category calculation unit that calculates the corresponding state category of the controlled object based on the generated state category map. The state category map generation unit classifies the operation data into and outside the attention area and generates state category maps corresponding to the inside of the attention area and the outside of the attention area, respectively. Other means will be described in the embodiments for carrying out the invention.

[0009] According to the present invention, it is possible to provide a management device, a management method, and a control system that appropriately improve the control characteristics of a controlled object.

[0010] This is a configuration diagram of a control system including a management device according to the first embodiment. This is a diagram showing an example of the hardware configuration of the management device according to the first embodiment. This is an explanatory diagram showing a specific example of a plant that is the control target of the management device according to the first embodiment. This is an explanatory diagram showing a specific example of a region of interest in the management device according to the first embodiment. This is a flowchart showing the processing flow of the management device according to the first embodiment. This is a flowchart showing the processing flow of generating a state category map in the management device according to the first embodiment. This is an explanatory diagram of the state category map within the region of interest in the management device according to the first embodiment. This is an explanatory diagram of the state category map outside the region of interest in the management device according to the first embodiment. This is a flowchart showing the processing flow of calculating and displaying state categories in the management device according to the first embodiment. This is an example of a display screen for state categories within the region of interest in the management device according to the first embodiment (1). This is an example of a display screen for state categories within the region of interest in the management device according to the first embodiment (2). This is an example of a display screen for state categories outside the region of interest in the management device according to the first embodiment. This is a configuration diagram of a control system including a management device according to the second embodiment. This is an example of a display screen for state category transitions and control operations in the management device according to the second embodiment (1). This is an example of a display screen for state category transitions and control operations in the management device according to the second embodiment (2). This is an example of a display screen showing state category transitions and control operations in the management device according to the second embodiment (part 3).

[0011] In the following, as an example of implementing the present invention, we will describe the case where the managed object (controlled object) of the management device 10 (see Figure 1) is a predetermined plant 20 (see Figure 1). In this embodiment, the predetermined plant 20 may be, for example, a power plant or a chemical plant, as well as an oil refinery, a steel plant, a food processing plant, a pharmaceutical plant, or a water treatment plant. Note that the managed object of the management device 10 is not limited to a plant, but may also be a robot, a vehicle, a ship, an aircraft, etc.

[0012] <Configuration of the control system of the first embodiment> Figure 1 is a configuration diagram of a control system 100 including a management device 10 according to the first embodiment. The control system 100 shown in Figure 1 is a system that controls the equipment of the plant 20 by the management device 10. As shown in Figure 1, the control system 100 includes a management device 10, a sensor 30, an input device 40, a display device 50, and a cloud 60, which are connected by wire or wireless.

[0013] In the following, we will briefly describe the plant 20, which is the target of control of the control device 10, as well as the sensor 30, input device 40, display device 50, etc., and then describe the control device 10 in detail.

[0014] Plant 20 is equipped with predetermined devices that serve as operating terminals, such as valves and heaters, as in a power plant. Sensor 30 acquires predetermined detection values ​​and environmental information from the devices in plant 20. Examples of such sensors 30 include cameras, distance sensors, radar, force sensors, temperature sensors, angle sensors, flow sensors, pressure sensors, voltage sensors, and current sensors.

[0015] The sensor 30 may be installed inside the plant 20 or outside the plant 20. In other words, the sensor 30 may be installed in either location. An example of the sensor 30 being installed outside the plant 20 is an ambient temperature sensor that detects the temperature of the surrounding environment of the plant 20. Furthermore, equipment for acquiring data such as power demand related to the control of the plant 20 from outside the plant 20 is also included in the sensor 30. In addition, soft sensors calculated from the detected values ​​of the sensor 30 are also included in the sensor 30.

[0016] The input device 40 is used by the operator M1 to input management information and commands to the control device 10. Examples of such input devices 40 include a keyboard, mouse, and joystick.

[0017] The display device 50 is used to present the calculation results of the control device 10 to the operator M1. For example, a liquid crystal display can be used as such a display device 50. Alternatively, a touch-panel mobile terminal that combines the functions of both the input device 40 and the display device 50, such as a smartphone or tablet, may be used as the input device 40 and the display device 50.

[0018] The cloud 60 has a server (not shown) that performs predetermined communication with the input device 40, the display device 50, and the management device 10 via a network. The processing results of the server are provided to the management device 10 via the network and are also displayed on the display device 50 as appropriate.

[0019] <Configuration of the Management Device> The management device 10 is a device that controls and manages the plant 20. Such a management device 10 can be a computer such as a personal computer, tablet, or smartphone. The management device 10 may also be configured by connecting multiple computers via communication lines or a network. Furthermore, the functions of the management device 10 may be distributed among multiple computers, such as a cloud server or edge server.

[0020] As shown in Figure 1, the management device 10 is configured to include a data acquisition unit 11, a storage unit 12, a communication unit 13, and a calculation unit 14. These data acquisition unit 11, storage unit 12, communication unit 13, and calculation unit 14 are connected via an internal bus (not shown).

[0021] Figure 2 shows an example of the hardware configuration of the management device 10. As shown in Figure 2, the management device 10 is configured with a processor 10a, RAM (Random Access Memory) 10b, ROM (Read Only Memory) 10c, storage 10d (HDD: Hard Disk Drive or SSD: Solid State Drive), a communication interface 10e, an input / output interface 10g, and a media interface 10h. These components are connected via an internal bus 10i in the management device 10.

[0022] The processor 10a shown in Figure 2 is the hardware of the arithmetic unit 14 (see Figure 1) of the management device 10. The RAM 10b, ROM 10c, and storage 10d are the hardware of the storage unit 12 (see Figure 1) of the management device 10. The processor 10a reads the programs stored in the ROM 10c and storage 10d and loads them into the RAM 10b, thereby executing various processes.

[0023] The communication interface 10e shown in Figure 2 is an interface for communication between the management device 10 and the equipment, sensors 30, and cloud 60 of the plant 20. The input / output interface 10g handles data input from the input device 40 and data output to the display device 50. Here, the communication interface 10e and the input / output interface 10g function as the communication unit 13 (see Figure 1) of the management device 10.

[0024] External recording media 70, such as an optical disc 70a or a USB memory 70b, are appropriately connected to the media interface 10h. The media interface 10h functions as the data acquisition unit 11 (see Figure 1) of the management device 10. Examples of such media interfaces 10h include an optical disc drive for playing the optical disc 70a and a USB interface for reading the USB memory 70b. Note that "USB (Universal Serial Bus)" is a registered trademark.

[0025] Let's return to Figure 1 and continue the explanation. The data acquisition unit 11 shown in Figure 1 acquires predetermined data from an external recording medium 70, such as an optical disc 70a or a USB memory 70b. For example, the data acquisition unit 11 acquires operation data, which includes time-series data related to the operation of the plant 20, and area of ​​interest information, which represents a predetermined area in the data space of the operation data. "Operation data" and "area of ​​interest" will be explained later.

[0026] The memory unit 12 stores various types of data. As shown in Figure 1, the memory unit 12 includes an operation data storage unit 12a, a focus area storage unit 12b, and a state category map storage unit 12c. Details of these will be described later, but in general terms, they have the following functions.

[0027] The operation data storage unit 12a stores the operation data of the plant 20. Here, "operation data" refers to time-series data (data for each day and date) related to the operation of the plant 20. The operation data includes, for example, time-series data for each item, such as observed values ​​from the sensor 30 and operation records of the plant 20 entered by operator M1 via the input device 40. Data acquired from the cloud 60 or external recording medium 70 may also be stored in the operation data storage unit 12a. The format of the data stored in the operation data storage unit 12a may include time-series data, tables and graphs, as well as calculation formulas that can calculate operation data, flowcharts, simulators, and documents.

[0028] As a result, the data acquisition unit 11 may acquire not only data acquired from the external recording medium 70, but also observed values ​​from the sensor 30, operation records of the plant 20 input via the input device 40, and data from the cloud 60 via the communication unit 13.

[0029] The focus area storage unit 12b stores focus area information for the control of the plant 20. Here, the "focus area" is an area in the operating data space that categorizes the state of the plant 20 on a physical scale in order to properly control the plant 20. For example, this includes the target area when controlling the state of the plant 20 to be within a predetermined range, and the prohibited area when controlling the state of the plant 20 to avoid being within a predetermined range.

[0030] In other words, the area of ​​interest is a predetermined area in the driving data space, and data such as upper and lower limits are stored in the area of ​​interest storage unit 12b in the form of tables, graphs, or calculation formulas. Furthermore, data entered by the operator M1 via the input device 40, data acquired from the cloud 60 or external recording medium 70, and calculation results from AI (Artificial Intelligence) may also be stored in the area of ​​interest storage unit 12b.

[0031] In other words, the region of interest is a region in the data space of the operating data that is defined by upper and lower limits of one or more dimensions of the operating data. The region of interest includes a target region and a prohibited region as predetermined regions. The target region is the region of the target state that indicates control to keep the state of the plant 20 within a predetermined range. The prohibited region is the region of the prohibited state that indicates control to prevent the state of the plant 20 from falling within the predetermined range.

[0032] Furthermore, a physical measure refers to the means of measuring a physical quantity, or the quantity obtained by that means. Therefore, physical measures include directly measurable quantities such as distance, weight, time, and volume, as well as quantities indirectly measured from other quantities such as temperature, density, and acceleration. The measure of a physical quantity may also be quantitative data. Thus, the area of ​​interest refers to a predetermined region within the driving data space classified by physical measures.

[0033] The state category map storage unit 12c stores the state category map, which is the result of the calculations performed by the state category map generation unit 14b. Here, the "state category map" is a representation that associates the state of the plant 20 (observed state values), which are observed as values ​​at each point in time in the operating data and represented as a continuous vector, with discrete categories (state categories). The data stored in the state category map storage unit 12c may be in the form of a table or graph, or it may be a decision tree or a calculation formula.

[0034] In other words, the state category map is a map that associates the state of plant 20 with a predetermined state category based on the acquired operating data of plant 20 and the information of the area of ​​interest. The predetermined category is a division of the state of plant 20 into categories, which will be described later, and each of these divided states is called a state category. The state category can represent the state of plant 20 using discrete categories. The predetermined category is, for example, the category of "steam temperature" or the category of "rate of change of steam temperature" as described later in Figure 7.

[0035] The communication unit 13 communicates with the management device 10, the plant 20, the sensors 30, as well as the input device 40, the display device 50, and the cloud 60. The communication unit 13 is connected to the storage unit 12 and the arithmetic unit 14 via an internal bus. As a result, data acquired from the plant 20, etc., via the communication unit 13 is stored in the storage unit 12. The calculation results of the arithmetic unit 14 are transmitted to the plant 20, etc., via the communication unit 13.

[0036] The calculation unit 14 generates data related to the control of the plant 20. Note that at least a portion of the calculation unit 14's processing may be performed by AI. As shown in Figure 1, the calculation unit 14 includes a focus area determination unit 14a, a state category map generation unit 14b, a state category calculation unit 14c, and a state category display unit 14d. Note that the functions of each functional unit of the calculation unit 14 may be replaced by the cloud 60.

[0037] Details of each function of the arithmetic unit 14 will be described later, but in general terms, it has the following functions.

[0038] In other words, the area of ​​interest determination unit 14a determines whether the values ​​of the operation data stored in the operation data storage unit 12a at each point in time, and the observed values ​​of the state of the plant 20 input via the communication unit 13, etc., are within the range of the area of ​​interest represented by the data stored in the area of ​​interest storage unit 12b. That is, the area of ​​interest determination unit 14a determines whether the state of the plant 20 is within the area of ​​interest based on the acquired area of ​​interest information. The determination result of the area of ​​interest determination unit 14a is input to the state category map generation unit 14b, the state category calculation unit 14c, and the state category display unit 14d.

[0039] The state category map generation unit 14b generates a state category map based on the operation data stored in the operation data storage unit 12a and the area of ​​interest information stored in the area of ​​interest storage unit 12b. That is, the state category map generation unit 14b generates a state category map that associates the state of the plant 20 with a predetermined state category based on the acquired operation data and area of ​​interest information. The state category map generated by the state category map generation unit 14b is stored in the state category map storage unit 12c.

[0040] The state category calculation unit 14c calculates a state category corresponding to the observed state of the plant 20, which is input via the communication unit 13, etc., based on the state category map stored in the state category map storage unit 12c and the determination result of the area of ​​interest determination unit 14a. In other words, the state category calculation unit 14c calculates the corresponding state category of the plant 20 based on the generated state category map. The calculation result of the state category calculation unit 14c is input to the state category display unit 14d.

[0041] The status category display unit 14d generates image information for displaying the calculation result of the status category calculation unit 14c on the display device 50, and outputs it to the display device 50. In other words, the status category display unit 14d displays the status category calculated by the status category calculation unit 14c on the display device 50. This allows operator M1 to easily grasp the status of the plant 20.

[0042] FIG. 3 is an explanatory diagram showing a specific example of a plant 20 that is a control target of the management device. In the example of FIG. 3, the case where the plant 20 is a thermal power generation plant is shown. In the plant 20 having the configuration shown in FIG. 3, a boiler drum 21, a first superheater 22, a second superheater 23, and a steam turbine 24 are sequentially connected via steam pipes.

[0043] The steam in the boiler drum 21 is superheated in the first superheater 22 and then further superheated in the downstream second superheater 23. The steam thus superheated in the second superheater 23 and having its temperature increased is led to the steam turbine 24. Then, the steam turbine 24 rotates by the wind pressure of the steam, and power generation is performed by the rotation of the rotor of a generator (not shown) integrally with the steam turbine 24.

[0044] As shown in FIG. 3, the downstream end of a spray pipe 25 is connected to the steam pipe between the first superheater 22 and the second superheater 23. A nozzle-shaped spray 26 is provided at the downstream end of the spray pipe 25. The spray pipe 25 is provided with an adjustment valve 27 for adjusting the injection amount of compressed water. Then, the compressed water injected through the spray 26 is mixed with the steam to lower the temperature of the steam. Note that the larger the amount of compressed water injected through the spray 26 (that is, the larger the opening degree of the adjustment valve 27), the larger the amount of decrease in the steam temperature.

[0045] Also, a temperature sensor 28a and a flow rate sensor 28b are provided in the steam pipe between the second superheater 23 and the steam turbine 24. The opening degree of the adjustment valve 27 is adjusted so that the detected value (steam temperature) of the temperature sensor 28a approaches a predetermined target temperature, and the steam flow rate is monitored by the flow rate sensor 28b. Note that the temperature sensor 28a and the flow rate sensor 28b correspond to the sensor 30 shown in FIG. 1.

[0046] Further, a pressure sensor 29a and a liquid level sensor 29b are provided in the boiler drum 21. Then, it is monitored whether the detected value of the pressure sensor 29a (the pressure of the steam) and the detected value of the liquid level sensor 29b (the steam) are within a safe range, and the combustion amount (not shown) of the boiler and the water supply amount to the drum (not shown) are appropriately adjusted. Note that the pressure sensor 29a and the liquid level sensor 29b also correspond to the sensor 30 shown in FIG. 1.

[0047] FIG. 4 is a specific example of a region of interest when controlling the plant 20 in the example of FIG. 3. In the example of FIG. 4, the case where the control target is to make the steam temperature static at 500 ± 0.5 ° C is shown. That is, the region of interest shown in FIG. 4 is the peripheral region including the region of 499.5 ≦ steam temperature [° C] ≦ 500.5, which is the target state of the plant 20, and the region of -0.5 ≦ change rate of steam temperature [° C / min] ≦ 0.5, and the region of 494.5 ≦ steam temperature [° C] ≦ 505.5 and the region of -2.5 ≦ change rate of steam temperature [° C / min] ≦ 2.5 (target region).

[0048] Further, a plurality of regions can be defined in the region of interest when controlling the plant 20 in the example of FIG. 3. For example, as a control constraint, when the steam pressure must not exceed 15 MPa, a region of 13 ≦ steam pressure [MPa] that includes the region of 15 ≦ steam pressure [MPa], which is the prohibited state of the plant 20, can be defined as one of the regions of interest simultaneously with the example of FIG. 4.

[0049] <Processing of the management device> FIG. 5 is a flowchart showing the processing flow of the management device 10 in FIG. 1 (see FIG. 1). Note that the series of processes shown in FIG. 5 is started by the operation of the operator M1 via the input device 40.

[0050] In step St1 of FIG. 5, the management device 10 generates a state category map by the region of interest determination unit 14a and the state category map generation unit 14b. The specific processing of step St1 will be described later. Note that in the present embodiment, it is characterized in that state category maps are generated for the regions inside and outside the region of interest.

[0051] Next, in step St2 of Figure 5, the control device 10 measures the state of the plant 20. Specifically, the control device 10 acquires the detected values ​​of the sensor 30, etc., via the communication unit 13 using the data acquisition unit 11.

[0052] Next, in step St3 of Figure 5, the control device 10 calculates a state category using the area of ​​interest determination unit 14a and the state category calculation unit 14c, and displays information about the state of the plant 20 on the display device 50 using the state category display unit 14d. The specific processing of step St3 will be described later. In this embodiment, the state category is calculated based on the corresponding state category map depending on whether the state of the plant 20 is inside or outside the area of ​​interest, and is presented to the operator M1 using an appropriate display method.

[0053] Next, in step St4 of Figure 5, the control device 10 determines whether or not to terminate the process. If the control device 10 terminates its process in step St4 (St4: Yes), the control device 10 terminates the series of processes (end). For example, when the operation of the plant 20 is terminated or when operation is interrupted for maintenance, the process in the flowchart of Figure 5 is terminated.

[0054] The trigger for the completion of processing by the control device 10 may be an operation of the input device 40 by an operator, or it may be based on the judgment of the AI.

[0055] On the other hand, if the management device 10 continues processing without terminating in step St4 (St4: No), the management device 10 proceeds to step St5.

[0056] In step St5, the control device 10 determines whether or not the area of ​​interest has been updated. The decision of whether or not to update the area of ​​interest is made by the plant operator or AI.

[0057] If the area of ​​interest is updated in step St5 (St5: Yes), the management device 10 returns to step St1. In this case, the state category map is regenerated based on the updated area of ​​interest.

[0058] On the other hand, if there is no update to the area of ​​interest in step St5 (St5: No), the control device 10 returns to step St2. In this case, the state category calculation unit 14c calculates the state category of the plant 20 based on the current area of ​​interest and the state category map.

[0059] <Generation of State Category Maps> Figure 6 is a flowchart relating to the generation of state category maps (see Figure 1). The flowchart in Figure 6 shows the details of the process in step St1 of Figure 5 (generation of state category maps). In other words, the flowchart in Figure 6 shows the series of processes by which the management device 10 generates state category maps for each region from the operation data stored in the operation data storage unit 12a and the area of ​​interest information stored in the area of ​​interest storage unit 12b, using the area of ​​interest determination unit 14a and the state category map generation unit 14b.

[0060] In step St11, the management device 10 classifies the operation data into areas of interest and areas of interest. More specifically, the management device 10 first reads the operation data stored in the operation data storage unit 12a and also reads the area of ​​interest information stored in the area of ​​interest storage unit 12b. The management device 10 also uses the area of ​​interest determination unit 14a to determine whether the value of the operation data at each point in time is inside or outside the area of ​​interest, and classifies the operation data into areas of interest and areas of interest. In this way, the state category map generation unit 14b classifies the operation data into areas of interest and areas of interest using the area of ​​interest determination unit 14a.

[0061] Next, in step St12, the management device 10 generates a state category map within the area of ​​interest from the area of ​​interest information stored in the area of ​​interest storage unit 12b using the state category map generation unit 14b, and stores it in the state category map storage unit 12c. In this case, the state category map generation unit 14b generates the state category map corresponding to the interior of the area of ​​interest by dividing it into state categories corresponding to predetermined divisions based on a physical scale.

[0062] Figure 7 is an explanatory diagram of the state category map within the region of interest. In the example in Figure 7, the region of interest information is as in the example in Figure 4, and state categories C1 to C55 are defined (classified) by a combination of "steam temperature" and "rate of change of steam temperature" (a predetermined classification) that defines the region of interest. Note that "steam temperature" and "rate of change of steam temperature" correspond to the dimensions of the operating data.

[0063] Specifically, the steam temperature range of 494.5°C to 505.5°C, which is the area of ​​interest, is divided into increments of 1°C. In addition, a range for the rate of change of steam temperature is set corresponding to each range of steam temperature.

[0064] In the example in Figure 7, the state where the "steam temperature" is between 494.5 [°C] and 495.5 [°C] and the "rate of change in steam temperature" is in the range of 1.5 [°C / min] to 2.5 [°C / min] is defined (classified) as state category C1. Similarly, other state categories C2, C3, ..., C55 are also defined (classified).

[0065] In the example shown in Figure 7, the division width for the "steam temperature" dimension is set to 1 [°C], and the division width for the "rate of change in steam temperature" dimension is set to 1 [°C / min], resulting in 55 state categories. However, this is just one example, and the division width for each dimension and the number of state categories can be set to any value. The number of state categories is set by the user via the input device 40 (see Figure 1).

[0066] Furthermore, in step St12 of Figure 6, if there are multiple regions of interest, a state category map is generated for each region of interest as described above and stored in the state category map storage unit 12c.

[0067] Next, in step St13 of Figure 6, the management device 10 uses the state category map generation unit 14b to generate a state category map for the area of ​​interest from the operating data outside the area of ​​interest classified in step St11, and stores it in the state category map storage unit 12c. In this case, the state category map generation unit 14b generates a state category map corresponding to the area outside the area of ​​interest based on the distance to a predetermined point in the data space.

[0068] Figure 8 is an explanatory diagram of the state category map outside the area of ​​interest. In the example in Figure 8, cluster centers are defined (classified) by the values ​​of each item of operating data (each item of the observed state of plant 20), such as steam temperature, rate of change of steam temperature, and steam pressure, and state categories C56 to C100 corresponding to each cluster center are defined (classified).

[0069] Specifically, a state that is geographically close to the cluster center, with a steam temperature of 480.1 [°C], a steam temperature change rate of 0.3 [°C / min], and a steam pressure of 0.13 [MPa], is defined (classified) as state category C56. Here, Euclidean distance can be used for the distance to the cluster center. Similarly, other state categories C57 to C100 are also defined (classified). In this way, the state category map generation unit 14b can generate a state category map corresponding to the area outside the region of interest based on the distance to the cluster center.

[0070] In step St13 of Figure 6, to calculate the cluster center in Figure 8, data clustering techniques are applied to the operating data outside the area of ​​interest classified in step St11. Here, the k-means method can be used as the data clustering technique.

[0071] In the example shown in Figure 8, the number of state categories is set to 45, but the number of state categories can be set to any value. Furthermore, the number of state categories can be set in advance by user operation via the input device 40 (see Figure 1), or the management device 10 may calculate it using data clustering technology (for example, the x-means method) that calculates the number of clusters from the data.

[0072] In this way, the management device 10 classifies the operating data into and outside the area of ​​interest, and generates the state category maps corresponding to the inside and outside of the area of ​​interest, respectively. Note that the order of steps St12 and St13 in Figure 6 may be reversed.

[0073] In this embodiment, the state category map generation unit 14b classifies the operation data inside and outside the area of ​​interest, and generates state category maps corresponding to the inside and outside of the area of ​​interest, thereby enabling discretization according to the area of ​​interest. In other words, the state category map generation unit 14b can set discretization according to target states and prohibited states inside the area of ​​interest, while outside the area of ​​interest, it can efficiently discretize with a predetermined number of states.

[0074] As a result, the control device 10 according to this embodiment can appropriately improve the control characteristics of the plant 20.

[0075] <Calculation and Display of State Categories> Figure 9 is a flowchart relating to the calculation and display of state categories (see Figure 5). The flowchart in Figure 9 shows the details of the process in step St3 of Figure 5 (calculation and display of state categories). In other words, the flowchart in Figure 9 shows the series of processes by which the management device 10 calculates the corresponding state category of the plant 20 from the state of the plant 20 measured in step St2 of Figure 5, the state category calculation unit 14c, and the state category display unit 14d, and displays it on the display device 50.

[0076] First, in step St31, the control device 10 uses the area of ​​interest determination unit 14a to determine whether the state of the plant 20 measured in step St2 of Figure 5 is inside or outside the area of ​​interest.

[0077] Next, in step St32, the management device 10 calculates the state category of the corresponding plant 20 using the state category calculation unit 14c. Then, in step St33, the management device 10 displays the state category of the corresponding plant 20 on the display device 50 using the state category display unit 14d.

[0078] The specific processing in steps St32 and St33 is switched depending on the determination result in step St31. In this embodiment, based on the determination result in step St31, the state category display unit 14d switches the display method of the state category depending on whether the state of the plant 20 is inside or outside the area of ​​interest.

[0079] In other words, if the result of the determination in step St31 is that the state of the plant 20 is inside the area of ​​interest (St31: Yes), the control device 10 calculates the state category within the area of ​​interest in step St32a. Then, in step St33a, the control device 10 displays the state category within the area of ​​interest on the display device 50.

[0080] To explain in more detail, first, in step St32a, the management device 10 reads the state category map stored in the state category map storage unit 12c that corresponds to the area of ​​interest, and calculates the corresponding state category of the plant 20 by comparing the range of values ​​defined in the state category map within the area of ​​interest (see Figure 7) with the state of the plant 20 measured in step St2 of Figure 5.

[0081] Next, in step St33a, the management device 10 displays the status category within the area of ​​interest.

[0082] Figures 10A and 10B show examples of displaying state categories within the area of ​​interest. In the examples in Figures 10A and 10B, the state categories of plant 20 are displayed on the display device 50 when the observed values ​​of the state of plant 20 are within the area of ​​interest as shown in the example in Figure 4.

[0083] As shown in Figure 10A, the state category display unit 14d displays the possible state categories within the area of ​​interest using a grid. In step St32a (see Figure 9), the state category display unit 14d displays the state category (current state) calculated from the observed state of the plant 20 using black circles, and also displays the target state for controlling the plant 20 using stars. This allows operator M1 to grasp the relationship between the current state and the target state of the plant 20 at a glance.

[0084] Furthermore, as shown in Figure 10A, the status category display unit 14d displays the current status of the plant 20 by creating a matrix display that associates each item of the operating data with the vertical and horizontal axes, respectively. In this way, the status category display unit 14d can present the observed values ​​of the multidimensional state of the plant 20 to the operator M1 in an easy-to-understand two-dimensional format.

[0085] Furthermore, as shown in Figure 10B, the state category display unit 14d displays the observed values ​​of the current state of the plant 20 and the range of values ​​for the target state in a table. This allows the operator M1 to easily confirm the detailed relationship between the current state and the target state of the plant 20.

[0086] Returning to Figure 9, if the result of the determination in step St31 indicates that the state of plant 20 is outside the area of ​​interest (St31: No), the control device 10 calculates the state category outside the area of ​​interest in step St32b. Then, in step St33b, the control device 10 displays the state category outside the area of ​​interest on the display device 50.

[0087] To explain in more detail, the management device 10 reads the state category maps stored in the state category map storage unit 12c that correspond to the area outside the area of ​​interest, calculates the distance between the cluster center defined in the state category map outside the area of ​​interest (see Figure 8) and the state of the plant 20 measured in step St2 of Figure 5, and calculates the state category corresponding to the cluster center with the shortest distance.

[0088] Next, in step St33b, the management device 10 displays a status category outside the area of ​​interest.

[0089] Figure 11 shows an example of displaying a state category outside the area of ​​interest. In the example in Figure 11, the state category of plant 20 is displayed on the display device 50 when the observed value of the state of plant 20 is outside the area of ​​interest.

[0090] As shown in Figure 11, the state category display unit 14d displays the representative state points (cluster centers) of state categories that can occur outside the area of ​​interest using white circles. In step St32b (see Figure 9), the state category display unit 14d displays the state category (current state) calculated from the observed state of the plant 20 using black circles. The state category display unit 14d also displays the representative point of the target state when controlling the plant 20 using a star, and the representative point of the prohibited state using a square.

[0091] In this embodiment, as an example, isometric mapping (Isomap), one of the data dimensionality reduction techniques, is used to display representative points of high-dimensional state categories in two dimensions. Isometric mapping is a technique that maps data to a lower dimension while maintaining the distance between data points in a high-dimensional space. This allows operator M1 to easily confirm the relationship between the current state of plant 20 (black circles) and the areas of interest (target areas (stars) and prohibited areas (squares)).

[0092] In step St33b, the state category display unit 14d may also display the observed values ​​of the current state of the plant 20 and the range of values ​​for the target state in a table, as shown in Figure 10B. This allows the operator M1 to easily confirm the detailed relationship between the current state and the target state of the plant 20.

[0093] <Configuration of the control system of the second embodiment> Figure 12 is a configuration diagram of the control system 101 including the management device 10 according to the second embodiment. The management device 10 of the second embodiment is configured by further including a control operation determination unit 14e, a control operation display unit 14f, and a control rule storage unit 12d in addition to the management device 10 of the first embodiment (see Figure 1).

[0094] In Figure 12, the control operation determination unit 14e and the control operation display unit 14f are located in the calculation unit 14. The control rule storage unit 12d is located in the storage unit 12.

[0095] The control operation determination unit 14e calculates the control operation for plant 20 from the current state category of plant 20 calculated by the state category calculation unit 14c and the control law stored in the control law storage unit 12d. The control operation display unit 14f then displays the control operation calculated by the control operation determination unit 14e on the display device 50.

[0096] Here, "control law" refers to the correspondence between the state categories of the plant 20, the optimal control variable, and the next state category to which the plant transitions when the optimal operation is performed. The data format for the control law may be a table or a calculation formula.

[0097] The control law is calculated by the management device 10 using machine learning techniques from the operation data stored in the operation data storage unit 12a. For example, conservative Q-learning, one of the offline reinforcement learning techniques, can be used to calculate the control law. Conservative Q-learning is effective in safely controlling the controlled object because it reduces risk by not overestimating the value of each operation of the controlled object.

[0098] In addition to input from the operator M1 via the input device 40, control rules may also be stored in the control rule storage unit 12d in a predetermined format, obtained from the cloud 60 or an external recording medium 70.

[0099] Figures 13A to 13C show examples of state category transitions and control operations for plant 20.

[0100] For example, in the example in Figure 13A, the relationship between the current state (black circle) and the target state (star) of plant 20 is displayed. In the example in Figure 13B, the path (arrow) to reach the target state (star) of plant 20 is shown on the display device 50. In the example in Figure 13C, the control path and control operations are shown on the display device 50.

[0101] As shown in Figure 13A, the control operation display unit 14f displays the current state (black circle) and the target state (star) of the plant 20 outside the area of ​​interest on the display device 50, and also displays the path to the target state (star) with an arrow.

[0102] Furthermore, as shown in Figure 13B, the control operation display unit 14f displays the path from outside the area of ​​interest (state category C59) to inside the area of ​​interest (state category C22) to the target state (star) using arrows.

[0103] Furthermore, as shown in Figure 13C, the control operation display unit 14f displays in a table the control path from the current state (black circle) to the target state (star), as well as the necessary control operations and operation quantities.

[0104] For example, in Figure 13A, the current state of plant 20 (indicated by a black circle) is state category C62, and the control path to reach the target state (indicated by a star) is to first transition from state category C62 (indicated by a black circle) to state category C59 (indicated by a white circle), and then transition from state category C59 (indicated by a white circle) to state category C22 (indicated by a star).

[0105] In this case, as shown in Figure 13C, it is indicated that the optimal operation to transition from the current state category C62 (black circle) to state category C59 (white circle) is to operate the control valve 27 to 80%. Note that the control valve 27 is an example and is not limited to this.

[0106] In this way, operator M1 can grasp the relationship between control operations and state transitions at a glance, and optimally control the plant 20 by inputting control operations to the plant 20 via the input device 40 according to the control operation display screen. Note that the control operations calculated by the control operation determination unit 14e are not limited to input by operator M1; for example, the control operations may be directly input to the plant 20 by AI.

[0107] <Effects> According to this embodiment, the control device 10 determines whether the state of the plant 20 is within the area of ​​interest based on the acquired area of ​​interest information using the area of ​​interest determination unit 14a. The control device 10 generates a state category map using the state category map generation unit 14b, which associates the state of the plant 20 with a predetermined state category based on the acquired operating data and the area of ​​interest information. In this case, the state category map generation unit 14b classifies the operating data into inside and outside the area of ​​interest and generates state category maps corresponding to the inside of the area of ​​interest and the outside of the area of ​​interest, respectively.

[0108] The state category map generation unit 14b generates state category maps corresponding to both the inside and outside of the region of interest, thereby enabling discretization according to the region of interest. In other words, within the region of interest, the state category map generation unit 14b can set discretization according to target states and prohibited states, while outside the region of interest, it can efficiently discretize with a predetermined number of states.

[0109] As a result, the control device 10 according to this embodiment can determine whether the state of the plant 20 is inside or outside the area of ​​interest using the area of ​​interest determination unit 14a, and perform operation control according to the state of the plant 20, thereby appropriately improving the control characteristics of the plant 20.

[0110] In particular, if the state of the plant 20 is within the region of interest, the state category calculation unit 14c calculates the state category using a state category map (see Figure 7) which is divided into state categories corresponding to predetermined divisions based on a physical scale.

[0111] As a result, even when the target state (star) and prohibited state (square) of the plant 20 are defined (categorized) by physical scale, the control device 10 clearly shows the relationship between the current state category and the target state (star) and prohibited state (square), allowing the operator M1 to accurately and easily grasp the state of the plant 20.

[0112] Furthermore, if the state of the plant 20 is outside the region of interest, the state category calculation unit 14c calculates a state category based on the distance to a predetermined point in the data space. Specifically, the state category calculation unit 14c calculates a state category using a cluster-centered state category map (see Figure 8). As a result, the management device 10 can efficiently represent the state of the plant 20 with a small number of state categories, even when the state of the plant 20 is represented by high-dimensional data.

[0113] Furthermore, the status category display unit 14d can switch the display of status categories (Figures 10A and 11) depending on whether the state of the plant 20 is inside or outside the area of ​​interest. This allows the status category display unit 14d to present the relationship between the current state of the plant 20 and the target state and prohibited state to the operator M1 with appropriate resolution. As a result, the operator M1 can properly monitor and operate the plant 20.

[0114] <<Modifications>> The management device 10 etc. related to this disclosure have been described above in the embodiments, but this disclosure is not limited to these descriptions and various modifications can be made.

[0115] For example, in this embodiment, we have described the case where there is only one controlled object, the plant 20. However, the number of controlled objects is not limited to one, and this embodiment can also be applied when there are multiple controlled objects.

[0116] Furthermore, although this embodiment describes a configuration in which the management device 10 includes a communication unit 13 (see Figure 1), it is not limited to this configuration. That is, the communication unit 13 may be omitted from the configuration in Figure 1 as appropriate, and the status category display unit 14d may transmit the display screen directly to the display device 50.

[0117] Furthermore, although this embodiment describes a case where the state category of the plant 20, which is the result of the state category calculation unit 14c, is displayed on the display device 50 by the state category display unit 14d, it is not limited to this. For example, the calculation result of the state category calculation unit 14c may be transmitted to an external server or a user's terminal device. This allows the management device 10 to provide the calculation result of the state category calculation unit 14c to an external party as data indicating the state category of the plant 20.

[0118] Furthermore, the management method executed by the management device 10 may be executed as a computer program. The aforementioned program can be provided via a communication line, or it can be written to a recording medium such as a CD-ROM and distributed.

[0119] Furthermore, this disclosure is not limited to the embodiments and includes various modifications. For example, the embodiments are described in detail for the purpose of clearly illustrating this disclosure and are not necessarily limited to having all the configurations described. Also, some of the configurations of the embodiments can be added, deleted, or replaced with other configurations.

[0120] Furthermore, each of the aforementioned configurations, functions, processing units, processing means, etc., may be implemented in hardware, either partially or entirely, by designing them as integrated circuits, for example. Alternatively, each of the aforementioned configurations, functions, etc., may be implemented in software by having the processor interpret and execute programs that realize each function. Information such as programs, tables, and files that realize each function can be stored in memory, a recording device such as a hard disk or SSD (Solid State Drive), or a recording medium such as an IC card, SD card, or DVD.

[0121] Furthermore, the control lines and information lines shown are those deemed necessary for explanation purposes, and do not necessarily represent all control lines and information lines in the actual product. In reality, it can be assumed that almost all components are interconnected.

[0122] 10 Management device 11 Data acquisition unit 12 Storage unit 12a Operation data storage unit 12b Area of ​​interest storage unit 12c State category map storage unit 12d Control law storage unit 13 Communication unit 14 Calculation unit 14a Area of ​​interest determination unit 14b State category map generation unit 14c State category calculation unit 14d State category display unit 14e Control operation determination unit 14f Control operation display unit 20 Plant (controlled object) 30 Sensor 40 Input device 50 Display device 60 Cloud 70 External recording medium 100 Control system

Claims

1. A control device comprising: a data acquisition unit that acquires operation data including time-series data relating to the operation of a controlled object, and focus area information representing a predetermined area in the data space of the operation data; a focus area determination unit that determines whether the state of the controlled object is within the focus area based on the acquired focus area information; a state category map generation unit that generates a state category map that associates the state of the controlled object with a state category indicating a predetermined division based on the acquired operation data and the focus area information; and a state category calculation unit that calculates the corresponding state category of the controlled object based on the generated state category map, wherein the state category map generation unit classifies the operation data into or outside the focus area and generates state category maps corresponding to the inside of the focus area and the outside of the focus area, respectively.

2. The management device according to claim 1, wherein the state category map generation unit generates the state category map corresponding to the interior of the region of interest by dividing it into state categories corresponding to predetermined divisions based on a physical scale, and generates the state category map corresponding to the exterior of the region of interest based on the distance to a predetermined point in the data space.

3. The management device according to claim 1, characterized in that the region of interest is a region in the data space of the operating data, where the predetermined region is indicated by the upper and lower limits of one or more dimensions of the operating data.

4. The control device according to claim 3, characterized in that the predetermined region is a region of a target state that indicates control to keep the state of the controlled object within a predetermined range.

5. The control device according to claim 3, characterized in that the predetermined area is a prohibited state area that indicates control to prevent the state of the controlled object from falling within a predetermined range.

6. The management device according to claim 1, further comprising a state category display unit for displaying the calculated state categories on a display device.

7. The management device according to claim 6, wherein the state category display unit switches the method of displaying the state category depending on whether the state of the controlled object is inside or outside the area of ​​interest.

8. The control device according to claim 1, characterized in that the state category represents the state of the controlled object in discrete categories.

9. A management method characterized by the following steps: a data acquisition unit acquires operation data including time-series data relating to the operation of a controlled object and attention area information representing a predetermined area in the data space of the operation data; a attention area determination unit determines whether the state of the controlled object is within the attention area based on the acquired attention area information; a state category map generation unit generates a state category map that associates the state of the controlled object with a predetermined classification of state categories based on the acquired operation data and the attention area information; a state category calculation unit calculates the corresponding state category of the controlled object based on the generated state category map; and the state category map generation unit further classifies the operation data into or outside the attention area and generates state category maps corresponding to the inside of the attention area and the outside of the attention area, respectively.

10. A control system comprising a controlled object and a management device for managing the controlled object, wherein the management device comprises: a data acquisition unit that acquires operation data including time-series data relating to the operation of the controlled object, attention area information representing a predetermined area in the data space of the operation data, and a control law including information on the optimal operation for each state of the controlled object; a attention area determination unit that determines whether the state of the controlled object is within the attention area based on the acquired attention area information; a state category map generation unit that generates a state category map that associates the state of the controlled object with a state category indicating a predetermined division based on the acquired operation data and the attention area information; a state category calculation unit that calculates the corresponding state category of the controlled object based on the generated state category map; and a control operation determination unit that determines a control operation of the controlled object based on the calculated state category and the acquired control law, wherein the controlled object comprises: a predetermined device as the operating end of the controlled object; and a sensor that acquires a predetermined detection value in the predetermined device, and the management device A control system comprising: a state category map generation unit that classifies the operating data into and out of a region of interest, and generates state category maps corresponding to the inside of the region of interest and the outside of the region of interest, respectively; a state category calculation unit that calculates the state category of the corresponding controlled object based on the detected value detected by the sensor, the acquired region of interest information, and each of the generated state category maps; and a control operation determination unit that determines a control operation for the predetermined equipment based on the calculated state category and the acquired control law.

Images